10 results on this page
A community developer created Qwen3.8-23B-Mini-Me by strategically removing layers from Qwen3.8-27B, reducing the model to approximately 22.7B parameters without severe reasoning degradation. The model is reported to work well for coding, agentic tasks, and multi-turn chats, but it has not yet been benchmarked and struggles more with edge cases and underspecified prompts.
Nvidia announced full-scale production of its Spectrum-X Ethernet Photonics co-packaged optical switch for AI factories, using 200G-per-lane links. SK hynix simultaneously published a CPO technology roadmap targeting over 100 Tb/s per node, sub-1 pJ/bit energy consumption, and sub-10 ns chip-to-chip latency.
Marvell and Google are deepening their partnership on custom AI chip development. Marvell is giving Google the right to purchase up to $12.2 billion in Marvell stock, strengthening their long-term alignment in the AI hardware race.
Waymo has developed a custom chip designed to improve the performance of its robotaxis. The chip also reduces the company’s reliance on third-party suppliers such as Nvidia.
Investor Duan Yongping bought back 301,400 Alibaba shares worth about $28.93 million, representing only 0.15% of his portfolio. The small position suggests renewed observation rather than strong conviction, especially as Alibaba explores AI cooperation with Apple while PDD Holdings receives a much larger allocation.
Apple Machine Learning presents a lexical-intervention approach for improving cross-lingual knowledge transfer when target-language data is scarce. The work targets downstream capabilities such as scientific reasoning, commonsense inference, and world knowledge without relying heavily on parallel data, translation systems, or auxiliary models.
Apple Machine Learning studies how to mix scarce target-domain data with abundant generic data during language-model pretraining. Across more than 2,000 training runs, the research examines the trade-off between insufficient target exposure, repeated examples, diminishing returns, and overfitting.
New releases from Chinese competitors are narrowing the gap with Silicon Valley’s AI products, including ChatGPT and Gemini. Lower pricing is helping these services gain market share, alongside offerings such as DeepSeek and Kimi agents.
AI chip startup Fractile is reportedly in advanced talks to secure a valuation above $6.5 billion. The discussions follow a supply agreement with Anthropic and would represent more than a sixfold increase from its May valuation.
Marvell Technology has granted Google the right to purchase up to $12.2 billion of Marvell shares as part of a custom-chip arrangement. Google will buy Marvell’s custom chips, and Marvell’s stock rose as much as 14% after the disclosure.